Files
marvinandClaude Opus 5.5 87dfcd55b7 Add resource limits, abuse protection and overload handling
Running games always continue; new work is refused when there is no
capacity; abuse is rejected immediately.

- Docker: CPU/memory/pids caps per container (overridable in .env);
  uvicorn message size, queue and concurrency limits
- Server abuse limits: connections total/per IP (close 1013), message
  rate per connection (token bucket, close 1008 on floods), lobby
  creation rate per IP
- Load monitor: event loop lag, CPU of the own cgroup quota and
  throttling, with hysteresis (normal/busy/overloaded, severe lag
  overloads immediately)
- Admission control: create_lobby and start_game refused while
  overloaded or MAX_ACTIVE_GAMES are running; server_status message
  broadcast on changes; /health reports load but stays ok
- AI bridge: timeout sent with requests, in-flight cap, circuit breaker
- AI service: bounded workers and queue ("overloaded"), drops expired
  requests, degrades to the heuristic agent when busy or too slow
- Website: busy notices, Create/Start disabled when not accepting
  games, slow retry when the server is full
- Tests for limiters, load levels, admission control and AI service
  limits; docs updated

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-27 12:51:46 +02:00

92 lines
2.4 KiB
Python

"""Messages exchanged with web clients over the ``/ws`` websocket.
Client -> server messages are validated with pydantic. Server -> client
messages are plain dicts:
- ``{"type": "welcome", "player_id", "token", "name"}``
- ``{"type": "state", "lobby": LobbyView | null, "game": GameView | null, "events": [...]}``
- ``{"type": "error", "message", "code"?}``
- ``{"type": "server_status", "load", "accepting_games"}``
"""
from __future__ import annotations
from typing import Annotated, Any, Literal
from pydantic import BaseModel, Field, StringConstraints, TypeAdapter
PlayerName = Annotated[str, StringConstraints(strip_whitespace=True, min_length=1, max_length=20)]
LobbyCode = Annotated[str, StringConstraints(strip_whitespace=True, to_upper=True, min_length=4, max_length=12)]
class Hello(BaseModel):
type: Literal["hello"]
# Token of a previous session, used to reconnect after a page reload.
token: str | None = None
class CreateLobby(BaseModel):
type: Literal["create_lobby"]
name: PlayerName
class JoinLobby(BaseModel):
type: Literal["join_lobby"]
code: LobbyCode
name: PlayerName
class LeaveLobby(BaseModel):
type: Literal["leave_lobby"]
class AddAI(BaseModel):
type: Literal["add_ai"]
class RemovePlayer(BaseModel):
type: Literal["remove_player"]
player_id: str
class SetOptions(BaseModel):
type: Literal["set_options"]
hidden_bids: bool
class StartGame(BaseModel):
type: Literal["start_game"]
class GameAction(BaseModel):
type: Literal["action"]
action: dict[str, Any]
class ReturnToLobby(BaseModel):
type: Literal["return_to_lobby"]
ClientMessage = Annotated[
Hello | CreateLobby | JoinLobby | LeaveLobby | AddAI | RemovePlayer
| SetOptions | StartGame | GameAction | ReturnToLobby,
Field(discriminator="type"),
]
client_message_adapter: TypeAdapter[ClientMessage] = TypeAdapter(ClientMessage)
# Machine readable error codes (the frontend reacts to them).
ERROR_BUSY = "busy" # server has no capacity for this right now
ERROR_RATE_LIMITED = "rate_limited" # client sends too many messages / creates too many lobbies
# Websocket close code "Try Again Later" (RFC 6455 registry), used when full.
CLOSE_TRY_AGAIN_LATER = 1013
def error(message: str, code: str | None = None) -> dict:
message_dict: dict = {"type": "error", "message": message}
if code:
message_dict["code"] = code
return message_dict